Gemma 4 models now available on Amazon Bedrock
Amazon Web Services has launched the Gemma 4 family of open-weight AI models from Google DeepMind on its Amazon Bedrock platform, giving developers access to three distinct variants designed for different use cases. The release includes Gemma 4 31B for reasoning and coding workloads with a 256K-token context window, Gemma 4 26B-A4B optimized for cost and latency-sensitive applications, and Gemma 4 E2B designed for low-latency interactive scenarios. All models feature mixture-of-experts and dense architectures with built-in reasoning capabilities, native function calling, and multimodal support for text, image, video, and audio inputs across more than 35 languages. AWS has implemented new infrastructure optimizations in Bedrock specifically for price performance with these models, including enhanced tool calling, structured output generation, reasoning capabilities, and response streaming to help developers build more reliable generative AI applications using open-source models.
Why It Matters
This launch expands enterprise access to competitive open-source AI models through AWS's managed platform, potentially reducing costs and vendor lock-in compared to proprietary models. The multimodal capabilities and variety of model sizes allow organizations to match specific technical requirements with appropriate compute costs, while the enhanced Bedrock infrastructure suggests AWS is investing heavily in optimizing performance for open-source AI workloads.
This summary is generated using AI analysis of the original press release. Always refer to the original source for complete details.